Abstract: Aiming at the operational management in the context of shared manufacturing,the dynamic pricing optimization problem of production orders on a shared manufacturing platform over a finite time horizon was investigated taking into account the dynamic and random nature of customer demand in a shared manufacturing environment.Based on the limited manufacturing capacity resources,a stochastic dynamic programming model was developed with the objective of maximizing the profit of the shared manufacturing platform.An approximate dynamic programming approach was employed to transform the dynamic programming into an equivalent linear programming.To address the high dimensionality of states and decisions in the linear programming,an affine functional approximation was constructed by utilizing the available time slots of manufacturing equipments as basis function,and a greedy strategy was applied as a constraint sampling method to sample the state and decision sets,efficiently reducing the number of constraints and obtaining order dynamic pricing strategy.Through numerical experiments with different parameter settings,the stability and validity of the order dynamic pricing strategy was verified.

Key words: shared manufacturing, dynamic pricing, approximate dynamic programming

摘要： 针对平台下的运营管理问题,结合共享制造环境下客户需求的动态性和随机性特征,研究有限期内共享制造平台生产订单的动态定价优化问题。基于有限的制造设备产能资源,以共享制造平台利润最大化为目标,建立随机动态规划模型。采用近似动态规划方法将原始动态规划问题转换为等价的线性规划,针对线性规划中状态和决策维数过高的问题,以制造设备可用时间点为基函数,构建仿射近似函数近似状态价值,并以贪婪策略作为约束采样方法采样状态和决策集合,有效地减少了约束数量,得到了订单动态定价策略。通过不同参数设置下的仿真实验,验证了订单动态定价策略的稳定性和有效性。

关键词: 共享制造, 动态定价, 近似动态规划